Open MatthewScholefield opened 7 years ago
It is MIT. I will add later.
In addition, you may also consider my other project kann. It is a little more complicated for MLP (see mlp.c) and autoencoder (see ae.c), but it is faster (due to mini-batching, SSE and multi-threading) and far more flexible. You will be able to construct more complex networks such as CNN and LSTM.
Thanks, I didn't realize there was a command line interface for kann as well :). I'll try and use that then. (And later when you get a chance, also remember to add a license to kann)
The header kann.h shows the license, but yes, I will add a license.txt file at some point.
Thanks!
I was looking at using this but I can't since I couldn't find any license on the repository. My 2 cents are that something less restrictive like
Apache 2
orzlib
lets others use the code without concerns over something likeGPLv3
that is considered viral.